Designing Social Inquiry

From Wikipedia, the free encyclopedia

Designing Social Inquiry: Scientific Inference in Qualitative Research is a 1994 book written by Gary King, Robert Keohane, and Sidney Verba(KKV) that lays out guidelines for conducting qualitative research. The central thesis of the book is that qualitative and quantitative research share the same "logic of inference" (p. 3). KKV, as it is known, is considered a major breakthrough in elucidating the ways by which quantitative principles could improve the rigor of qualitative research.

According to KKV, a strong research design requires both qualitative and quantitative research, a research question that poses an important and real question that will contribute to the base of knowledge about this particular subject, and a literature review that develops this question through at least twenty years of literature, from which hypothesis (theory-driven) are then drawn. Data that is collected should be operationalized so that the next researcher can come along and develop the same study and achieve similar results. While gathering data the researcher should consider the observable implications of the theory in an effort to explain as much of the data as possible. This is in addition to examining the causal mechanisms that connect one variable to another.

The authors state that social scientists will avoid using causation rhetoric. This avoid the mistake of correlation does not equal causation. However to avoid causation statements can leave a person's work unfinished, unsatisfying, or less explained. One can "be bold" in stating causation but should also state the uncertainty in the inference. The authors state a researcher can make these bold, causation statements while allowing for variance by using what they call "variance of causal effect." For instance, in comparing democracies and stability, the authors discuss a parliamentary system and a presidential system. A president can have a high stabilizing effect, but can also be unstable depending on the country we're discussing. Thus, a presidential system has high variance. For some countries, this variance of causal effect is too high and a proportional representation system should be considered. In other words, in a study, one may place a number as to which if the variance reaches it, the theory should not be applied. This is a nice feature for it allows for falsifiability--a notion every scientist enjoys.